Health-system drivers of outpatient antibiotic use: evidence from prescribing, centralized supply, and catchment populations in public tertiary hospitals in Bihar, India

Background In low- and middle-income countries (LMICs), antibiotic use is shaped not only by clinical decision-making but also by health-system factors such as centralized drug procurement, supply availability, and population catchment pressures. However, empirical evidence linking outpatient prescribing patterns with state-level antibiotic supply and catchment characteristics remains limited. Bihar, one of India’s most populous and resource-constrained states, represents a critical yet under-studied setting for understanding these interactions. Methods We conducted a multicentre observational study across three public tertiary care hospitals in Bihar, SKMCH Muzaffarpur, DMCH Darbhanga, and JLNMCH Bhagalpur, using outpatient prescription data collected between 2023 and 2025 (n = 8,822). These data were triangulated with state-level antibiotic supply records from the Bihar Medical Services and Infrastructure Corporation Limited (BMSICL) for 2022–2025 and secondary data describing hospital catchment populations. Prescribing indicators, WHO AWaRe classification, essential drug list (EDL) status, and supply–prescription concordance were analysed to assess alignment with antimicrobial stewardship goals. Results Across all sites, outpatient antibiotic use was characterized by high polypharmacy (4–4.7 drugs per prescription), extensive empirical prescribing, and minimal use of culture-based diagnostics (≤ 2.6%). Broad-spectrum antibiotics dominated prescribing, with Amoxicillin–Clavulanic acid (42.8%), Azithromycin (13.1%), and Cefixime (6.9%) being most frequently used. While Access-category antibiotic use exceeded 50% at two sites, Watch-category use was disproportionately high at JLNMCH (58%), a hospital serving an extensive rural catchment with long patient travel times. Supply–prescription analysis showed strong alignment for commonly used antibiotics, reflecting stable state-level procurement patterns; however, continued supply and use of Watch-category and non-EDL antibiotics, including Moxifloxacin, raised stewardship concerns. Notably, irrational fixed-dose combinations were prescribed despite not being supplied, indicating parallel private-sector access. Generic prescribing exceeded 97% across sites, reflecting positive institutional norms. Conclusions Antibiotic prescribing in Bihar’s public tertiary hospitals is shaped by the interaction of centralized supply stability, diagnostic constraints, and catchment-level pressures rather than prescriber behaviour alone. Persistent availability of broad-spectrum antibiotics through state procurement appears to reinforce entrenched prescribing patterns, particularly in high-volume, rural-serving facilities. Effective antimicrobial stewardship in such settings requires context-sensitive strategies that integrate supply-chain reform, strengthened diagnostic pathways, and locally tailored prescriber support, moving beyond uniform, guideline-only approaches.